1
$\begingroup$

The person posting this answer appears to be this person, an astrophysicist with a Wikipedia page, an asteroid, and a long-established, personally curated and maintained list of launches.

They'd originally signed the post, but this was removed as part of normal SE housekeeping. There is no verifying information in the profile, where it should have appeared.

On one hand, this is not a good SE answer, because the supporting information is essentially of the form "it's somewhere in this list that I also wrote." There are some active users here that over time have established identity and authority in certain areas, but this it the first post by this user id. (spot checks did not seem to suggest this list contains verification either.)

On the other hand, there's a good chance the post is by who it seems to be by, and the answer has accumulated a score of up votes, and up voting is supposed to be one indicator of an answer's quality and possibly correctness, and if the answer were glaringly wrong I have a hunch other readers would have caught the problem quickly and pointed it out. At least I assume some of the up voters checked... right?

So, do user ids that appear to be people with Wikipedia pages and asteroids get a pass on standard procedure, or should I hold out and not accept until the answer gets some independent verification/validation, or the user is somehow verified, or one of the many people who double-checked before reflexively up voting takes a moment to edit the question and help the busy poster out?

$\endgroup$
  • $\begingroup$ Many problems in mathematics and engineering have special cases and singularities. I guess I'll consider this one classified as such, or an 'exception that proves the rule', whatever that means. Answer accepted. $\endgroup$ – uhoh Apr 22 '17 at 9:59
1
$\begingroup$

You wanted a more specific answer than what Rory gave you, so here is my best shot.

Sometimes we treat experience and prior history of good answers as a proxy for verification (like with Mark Adler, as you mentioned in the comments). This is not because these are good measures of verification in and of themselves, but because we recognize that there are things that users like Mark Adler can personally verify that would take an inordinate amount of resources to independently verify. We do this because Mark continues to be an active user in this community and there are other users who can challenge in cases where his information may be off. For all of these cases, we should always be cautious to validate what we can.

In the case of a new or unregistered user, even if you believe they might be a notable individual with relevant experience, I do not believe we should take their experience as a proxy. There is no guarantee of return on this risk.

It is important to identify, as Rory said, that the actual criteria for upvote and acceptance of an answer is its truth and helpfulness, not any factor of the poster. Usually, we require citations to validate an answer, and a citation can also make an answer more helpful. If an answer has not helped you, don't upvote or accept it. We have plenty of cases where other users can validate the answer and do upvote it to the top, despite the OP being unconvinced. The system still works to deliver good content to viewers. If people are not upvoting for good reasons, we have comments, flags, and moderators to intervene to get these answers fixed.

We always have to accept, though, that we cannot guarantee the truth of any answer.

$\endgroup$
  • $\begingroup$ Thank you for taking the time to work through this step-by-step. I'm not a fan of absolutes. Some answers validity are self evident and don't need links, and from time to time a users N excellent answers will convince me the N+1 th answer is right, especially if it also rings of truth, like this one for example. In this case I did something I very rarely do and bucked to "perceived peer pressure" and quit asking for links and just accepted. The problem here is that the data may be secret or at least difficult to verify. $\endgroup$ – uhoh May 8 '17 at 15:10
  • $\begingroup$ The problem is that nobody asked for verification, nobody said this is not a valid answer, and literally dozens of people up voted. So I figured that this may be one of those ambiguous situations where we run up against sensitive issues. So I sort-of just let it go. Usually I won't accept an answer unless I feel it's been answered to my satisfaction. But sometimes I'll accept it if I think the answer is basically good and seems to be useful to other people, and I can always ask another question later. $\endgroup$ – uhoh May 8 '17 at 15:16
  • $\begingroup$ I'll note that in the case where I stuck to my principles, ironically @RoryAlsop dug into me pretty hard for being impolite for challenging an answer. I'm just not going to pay attention to that user's opinion of how I or anyone else should think. But in the present case, I think I'll un-accept the answer to this question and pursue some additional facts. $\endgroup$ – uhoh May 8 '17 at 15:18
  • $\begingroup$ @uhoh The accepted answer checkmark isn't important enough for you to worry about un-accepting in this case, but I'm liberating you from feeling the need to accept an answer in the future due to peer pressure. $\endgroup$ – called2voyage May 8 '17 at 15:18
  • 1
    $\begingroup$ @uhoh "I'll note that in the case where I stuck to my principles, ironically \@RoryAlsop dug into me pretty hard for being impolite for challenging an answer." Now, there is a separate issue in how much you should pester a user to improve their answer. One or two comments should be enough, and then if they won't fix it--let it go. $\endgroup$ – called2voyage May 8 '17 at 15:19
  • $\begingroup$ One or two is a good average. But like I say I'm not a fan of absolutes. OK I'll leave the checkmark there. I have a hunch the answer is likely reasonable, and if anyone wants to dig into the linked list there, they can read further. OK thanks! $\endgroup$ – uhoh May 8 '17 at 15:23
  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ – called2voyage May 8 '17 at 19:10
1
$\begingroup$

If one or more answers helped you, and you want to accept one, you should accept the answer that most helped you.

Whilst knowing the background or qualifications of an answerer could help you with assurance over accuracy, the vast majority of accepts have nothing to do with that.

Remember if a better answer comes later you can even change your acceptance.

$\endgroup$
  • 1
    $\begingroup$ Let us continue this discussion in chat. $\endgroup$ – Rory Alsop Apr 21 '17 at 14:49
  • $\begingroup$ I'll just summarize, I'm asking specifically about a single case where cited independent verification is absent, and the weight and proof of the answer relies solely on the person's reputation, which happens from time to time in special cases. Can you constrain your answer to those parameters? $\endgroup$ – uhoh Apr 21 '17 at 15:01
  • 5
    $\begingroup$ No. And I don't in general approve of that as an approach. Anyone, no matter how experienced, could give a terrible/wrong answer. Should you upvote it without checking because it is from an expert? Definitely not! Sure it can affect your opinion, but as I said: upvote the post, not the poster. Of course you can do what you like- they are your votes :-) $\endgroup$ – Rory Alsop Apr 21 '17 at 15:09

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .